arXiv subjects

Bing Li

Publications and source records attributed to Bing Li.

At least 109 records · Page 6Linked to original sources

arXiv2026

Flash annealing-engineered wafer-scale relaxor antiferroelectrics for enhanced energy storage performance

Dielectric capacitors are essential for energy storage systems due to their high-power density and fast operation speed. However, optimizing energy storage density with concurrent thermal stability remains a substantial challenge. Here, we develop a flash annealing process with ultrafast heating and cooling rates of 1000 oC/s, which facilitates the rapid crystallization of PbZrO3 film within a mere second, while locking its high-temperature microstructure to room temperature. This produces compact films with sub-grain boundaries fraction of 36%, nanodomains of several nanometers, and negligible lead volatilization. These contribute to relaxor antiferroelectric film with a high breakdown strength (4800 kV/cm) and large polarization (70 uC/cm2). Consequently, we have achieved a high energy storage density of 63.5 J/cm3 and outstanding thermal stability with performance degradation less than 3% up to 250 oC. Our approach is extendable to ferroelectrics like Pb(Zr0.52Ti0.48)O3 and on wafer scale, providing on-chip nonlinear dielectric energy storage solutions with industrial scalability.

cond-mat.mtrl-sci↗

arXiv2026

Toward Inherently Robust VLMs Against Visual Perception Attacks

Autonomous vehicles rely on deep neural networks (DNNs) for traffic sign recognition, lane centering, and vehicle detection, yet these models are vulnerable to attacks that induce misclassification and threaten safety. Existing defenses (e.g., adversarial training) often fail to generalize and degrade clean accuracy. We introduce Vehicle Vision-Language Models (V2LMs), fine-tuned vision-language models specialized for autonomous vehicle perception, and show that they are inherently more robust to unseen attacks without adversarial training, maintaining substantially higher adversarial accuracy than conventional DNNs. We study two deployments: Solo (task-specific V2LMs) and Tandem (a single V2LM for all three tasks). Under attacks, DNNs drop 33-74%, whereas V2LMs decline by under 8% on average. Tandem achieves comparable robustness to Solo while being more memory-efficient. We also explore integrating V2LMs in parallel with existing perception stacks to enhance resilience. Our results suggest V2LMs are a promising path toward secure, robust AV perception.

cs.CV↗

arXiv2026

Multi-View Adaptive Contrastive Learning for Information Retrieval Based Fault Localization

Most studies focused on information retrieval-based techniques for fault localization, which built representations for bug reports and source code files and matched their semantic vectors through similarity measurement. However, such approaches often ignore some useful information that might help improve localization performance, such as 1) the interaction relationship between bug reports and source code files; 2) the similarity relationship between bug reports; and 3) the co-citation relationship between source code files. In this paper, we propose a novel approach named Multi-View Adaptive Contrastive Learning for Information Retrieval Fault Localization (MACL-IRFL) to learn the above-mentioned relationships for software fault localization. Specifically, we first generate data augmentations from report-code interaction view, report-report similarity view and code-code co-citation view separately, and adopt graph neural network to aggregate the information of bug reports or source code files from the three views in the embedding process. Moreover, we perform contrastive learning across these views. Our design of contrastive learning task will force the bug report representations to encode information shared by report-report and report-code views,and the source code file representations shared by code-code and report-code views, thereby alleviating the noise from auxiliary information. Finally, to evaluate the performance of our approach, we conduct extensive experiments on five open-source Java projects. The results show that our model can improve over the best baseline up to 28.93%, 25.57% and 20.35% on Accuracy@1, MAP and MRR, respectively.

cs.SE↗

arXiv2026

EXaMCaP: Subset Selection with Entropy Gain Maximization for Probing Capability Gains of Large Chart Understanding Training Sets

Recent works focus on synthesizing Chart Understanding (ChartU) training sets to inject advanced chart knowledge into Multimodal Large Language Models (MLLMs), where the sufficiency of the knowledge is typically verified by quantifying capability gains via the fine-tune-then-evaluate paradigm. However, full-set fine-tuning MLLMs to assess such gains incurs significant time costs, hindering the iterative refinement cycles of the ChartU dataset. Reviewing the ChartU dataset synthesis and data selection domains, we find that subsets can potentially probe the MLLMs' capability gains from full-set fine-tuning. Given that data diversity is vital for boosting MLLMs' performance and entropy reflects this feature, we propose EXaMCaP, which uses entropy gain maximization to select a subset. To obtain a high-diversity subset, EXaMCaP chooses the maximum-entropy subset from the large ChartU dataset. As enumerating all possible subsets is impractical, EXaMCaP iteratively selects samples to maximize the gain in set entropy relative to the current set, approximating the maximum-entropy subset of the full dataset. Experiments show that EXaMCaP outperforms baselines in probing the capability gains of the ChartU training set, along with its strong effectiveness across diverse subset sizes and compatibility with various MLLM architectures.

cs.LG↗

arXiv2026

QVLA: Not All Channels Are Equal in Vision-Language-Action Model's Quantization

The advent of Vision-Language-Action (VLA) models represents a significant leap for embodied intelligence, yet their immense computational demands critically hinder deployment on resource-constrained robotic platforms. Intuitively, low-bit quantization is a prevalent and preferred technique for large-scale model compression. However, we find that a systematic analysis of VLA model's quantization is fundamentally lacking. We argue that naively applying uniform-bit quantization from Large Language Models (LLMs) to robotics is flawed, as these methods prioritize passive data fidelity while ignoring how minor action deviations compound into catastrophic task failures. To bridge this gap, we introduce QVLA, the first action-centric quantization framework specifically designed for embodied control. In a sharp departure from the rigid, uniform-bit quantization of LLM-based methods, QVLA introduces a highly granular, channel-wise bit allocation strategy. Its core mechanism is to directly measure the final action-space sensitivity when quantizing each individual channel to various bit-widths. This process yields a precise, per-channel importance metric that guides a global optimization, which elegantly unifies quantization and pruning (0-bit) into a single, cohesive framework. Extensive evaluations on different baselines demonstrate the superiority of our approach. In the LIBERO, the quantization version of OpenVLA-OFT with our method requires only 29.2% of the original model's VRAM while maintaining 98.9% of its original performance and achieving a 1.49x speedup. This translates to a 22.6% performance improvement over the LLM-derived method SmoothQuant. Our work establishes a new, principled foundation for compressing VLA models in robotics, paving the way for deploying powerful, large-scale models on real-world hardware. Code will be released.

cs.CV↗

arXiv2026

ATLAS MDT TDC Simulations for LHC Run3 and HL-LHC

The Large Hadron Collider (LHC) started the Run 3 operation in 2022, and the peak instantaneous luminosity in Run 3 may reach 3 x 10^34 cm-2s-1. The ATLAS Monitored Drift Tube (MDT) chambers are the main component of the precision tracking system in the ATLAS muon spectrometer. It is important to understand any potential issues with the MDT Front-End (FE) readout electronics for an expected level-1 (L1) trigger rate of 100 kHz and a complex deadtime of over 5% for Run 3 operations. We use raw data collected in 2022 to emulate the expected hit rates in MDT chambers and perform a realistic simulation on the ATLAS Muon TDC (Time-to-Digital Converter) (AMT) chip with the current configuration. We study the AMT chip performances by analyzing the trigger/L1/readout buffer occupancies and hit loss fractions under different luminosities with L1 rate of 100 kHz by using the Modelsim software. The hit loss fraction of the hottest MDT chamber (BIL3C05) is lower than 5% due to FE readout, even at a luminosity of 5.01 x 10^34 cm-2s-1 with a deadtime of 5% and a L1 rate of 100 kHz, indicating that AMT can operate under Run 3 conditions without problems. The MDT trigger and readout electronics will be replaced for triggerless readout during High-Luminosity LHC (HL-LHC) runs. We also simulate the AMT behavior in the triggerless mode up to 7.44 x 10^34 cm-2s-1 and propose possible AMT configurations in case some FE electronics could not be replaced during the long shutdown 3 (LS3).

physics.ins-det↗

arXiv2026

Late Breaking Results: Conversion of Neural Networks into Logic Flows for Edge Computing

Neural networks have been successfully applied in various resource-constrained edge devices, where usually central processing units (CPUs) instead of graphics processing units exist due to limited power availability. State-of-the-art research still focuses on efficiently executing enormous numbers of multiply-accumulate (MAC) operations. However, CPUs themselves are not good at executing such mathematical operations on a large scale, since they are more suited to execute control flow logic, i.e., computer algorithms. To enhance the computation efficiency of neural networks on CPUs, in this paper, we propose to convert them into logic flows for execution. Specifically, neural networks are first converted into equivalent decision trees, from which decision paths with constant leaves are then selected and compressed into logic flows. Such logic flows consist of if and else structures and a reduced number of MAC operations. Experimental results demonstrate that the latency can be reduced by up to 14.9 % on a simulated RISC-V CPU without any accuracy degradation. The code is open source at https://github.com/TUDa-HWAI/NN2Logic

cs.LG↗

arXiv2026

Token Caching for Diffusion Transformer Acceleration

Diffusion transformers have gained substantial interest in diffusion generative modeling due to their outstanding performance. However, their computational demands, particularly the quadratic complexity of attention mechanisms and multi-step inference processes, present substantial bottlenecks that limit their practical applications. To address these challenges, we propose TokenCache, a novel acceleration method that leverages the token-based multi-block architecture of transformers to reduce redundant computations. TokenCache tackles three critical questions: (1) Which tokens should be pruned and reused by the caching mechanism to eliminate redundancy? (2) Which blocks should be targeted for efficient caching? (3) At which time steps should caching be applied to balance speed and quality? In response to these challenges, TokenCache introduces a Cache Predictor that hierarchically addresses these issues by (1) Token pruning: assigning importance scores to each token to determine which tokens to prune and reuse; (2) Block selection: allocating pruning ratio to each block to adaptively select blocks for caching; (3) Temporal Scheduling: deciding at which time steps to apply caching strategies. Experimental results across various models demonstrate that TokenCache achieves an effective trade-off between generation quality and inference speed for diffusion transformers.

cs.LG↗

arXiv2026

Learning Functional Graphs with Nonlinear Sufficient Dimension Reduction

Functional graphical models have undergone extensive development during the recent years, leading to a variety models such as the functional Gaussian graphical model, the functional copula Gaussian graphical model, the functional Bayesian graphical model, the nonparametric functional additive graphical model, and the conditional functional graphical model. These models rely either on some parametric form of distributions on random functions, or on additive conditional independence, a criterion that is different from probabilistic conditional independence. In this paper we introduce a nonparametric functional graphical model based on functional sufficient dimension reduction. Our method not only relaxes the Gaussian or copula Gaussian assumptions, but also enhances estimation accuracy by avoiding the ``curse of dimensionality''. Moreover, it retains the probabilistic conditional independence as the criterion to determine the absence of edges. By doing simulation study and analysis of the f-MRI dataset, we demonstrate the advantages of our method.

stat.ME↗

arXiv2026

FastFI: Enhancing API Call-Site Robustness in Microservice-Based Systems with Fault Injection

Fault injection is a key technique for assessing software reliability, enabling proactive detection of system defects before they manifest in production. However, the increasing complexity of microservice architectures leads to exponential growth in the fault-injection space, rendering traditional random injection inefficient. Recent lineage-driven approaches mitigate this problem through heuristic pruning, but they face two limitations. First, combinatorial-fault discovery remains bottlenecked by general-purpose SAT solvers, which fail to exploit the monotone and low-overlap structure of derived CNF formulas and typically rely on a static upper bound on fault size. Second, existing techniques provide limited post-injection guidance beyond reporting detected faults. To address these challenges, we propose FastFI, a fault-injection-guided framework to enhance the robustness of API call sites in microservice-based systems. FastFI features a DFS-based solver with dynamic fault injection to discover all valid combinatorial faults, and it leverages fault-injection results to identify critical APIs whose call sites should be hardened for robustness. Experiments on four representative microservice benchmarks show that FastFI reduces end-to-end fault-injection time by an average of 76.12\% compared to state-of-the-art baselines while maintaining acceptable resource overhead. Moreover, FastFI accurately identifies high-impact APIs and provides actionable guidance for call-site hardening.

cs.SE↗

arXiv2026

SoLA-Vision: Fine-grained Layer-wise Linear Softmax Hybrid Attention

Standard softmax self-attention excels in vision tasks but incurs quadratic complexity O(N^2), limiting high-resolution deployment. Linear attention reduces the cost to O(N), yet its compressed state representations can impair modeling capacity and accuracy. We present an analytical study that contrasts linear and softmax attention for visual representation learning from a layer-stacking perspective. We further conduct systematic experiments on layer-wise hybridization patterns of linear and softmax attention. Our results show that, compared with rigid intra-block hybrid designs, fine-grained layer-wise hybridization can match or surpass performance while requiring fewer softmax layers. Building on these findings, we propose SoLA-Vision (Softmax-Linear Attention Vision), a flexible layer-wise hybrid attention backbone that enables fine-grained control over how linear and softmax attention are integrated. By strategically inserting a small number of global softmax layers, SoLA-Vision achieves a strong trade-off between accuracy and computational cost. On ImageNet-1K, SoLA-Vision outperforms purely linear and other hybrid attention models. On dense prediction tasks, it consistently surpasses strong baselines by a considerable margin. Code will be released.

cs.CV↗

arXiv2026

Brightest GRB flare observed in GRB 221009A: bridge the last gap between flare and prompt emission in GRB

Flares are usually observed during the afterglow phase of Gamma-Ray Bursts (GRBs) in soft X-ray, optical and radio bands, but rarely in gamma-ray band. Despite the extraordinary brightness, GECAM-C has accurately measured both the bright prompt emission and flare emission of GRB 221009A without instrumental effects, offering a good opportunity to study the relation between them. In this work, we present a comprehensive analysis of flare emission of GRB 221009A, which is composed of a series of flares. Among them, we identify an exceptionally bright flare with a record-breaking isotropic energy $E_{\rm iso} = 1.82 \times 10^{53}$ erg of GRB flares. It exhibits the highest peak energy ever detected in GRB flares, $E_{\rm peak} \sim 300$ keV, making it a genuine gamma-ray flare. It also shows rapid rise and decay timescales, significantly shorter than those of typical X-ray flares observed in soft X-ray or optical band, but comparable to those observed in prompt emissions. Despite these exceptional properties, the flare shares several common properties with typical GRB flares. We note that this is the first observation of a GRB flare in the keV-MeV band with sufficiently high temporal resolution and high statistics, which bridges the last gap between prompt emission and flare.

astro-ph.HE↗

arXiv2025

Doctor Sun: A Bilingual Multimodal Large Language Model for Biomedical AI

Large multimodal models (LMMs) have demonstrated significant potential in providing innovative solutions for various biomedical tasks, including pathology analysis, radiology report generation, and biomedical assistance. However, the existing multimodal biomedical AI is typically based on foundation LLMs, thus hindering the understanding of intricate medical concepts with limited medical training data. Moreover, recent LLaVA-induced medical LMMs struggle to effectively capture the intricate relationship between the texts and the images. Therefore, we introduce Doctor Sun, a large multimodal generative model specialized in medicine, developed to encode, integrate, and interpret diverse biomedical data modalities such as text and images. In particular, Doctor Sun integrates a pre-trained vision encoder with a medical LLM and conducts two-stage training on various medical datasets, focusing on feature alignment and instruction tuning. Moreover, we release SunMed-VL, a wide-range bilingual medical multimodal dataset, along with all associated models, code, and resources, to freely support the advancement of biomedical multimodal research.

cs.LG↗

arXiv2025

Structure-Preserving Nonlinear Sufficient Dimension Reduction for Tensors

We introduce two nonlinear sufficient dimension reduction methods for regressions with tensor-valued predictors. Our goal is two-fold: the first is to preserve the tensor structure when performing dimension reduction, particularly the meaning of the tensor modes, for improved interpretation; the second is to substantially reduce the number of parameters in dimension reduction, thereby achieving model parsimony and enhancing estimation accuracy. Our two tensor dimension reduction methods echo the two commonly used tensor decomposition mechanisms: one is the Tucker decomposition, which reduces a larger tensor to a smaller one; the other is the CP-decomposition, which represents an arbitrary tensor as a sequence of rank-one tensors. We developed the Fisher consistency of our methods at the population level and established their consistency and convergence rates. Both methods are easy to implement numerically: the Tucker-form can be implemented through a sequence of least-squares steps, and the CP-form can be implemented through a sequence of singular value decompositions. We investigated the finite-sample performance of our methods and showed substantial improvement in accuracy over existing methods in simulations and two data applications.

math.ST↗

arXiv2025

Intersecting well approximable and missing digit sets

Let $b\geq3$ be an integer and $C(b,D)$ be the set of real numbers in $[0,1]$ whose $b$-ary expansion consists of digits restricted to a given set $D\subseteq\{0,\ldots,b-1\}$. Given an integer $t\geq2$ and a real, positive function $ψ$, let $W_{t}(ψ)$ denote the set of $x$ in $[0,1]$ for which $|x-p/t^{n}|<ψ(n)$ for infinitely many $(p,n)\in\mathbb{Z}\times\mathbb{N}$. We prove a general Hausdorff dimension result concerning the intersection of $W_{t}(ψ)$ with an arbitrary self similar set which implies that $\dim_{\rm H}(W_{t}(ψ)\cap C(b,D))\le\dim_{\rm H}W_{t}(ψ)\times \dim_{\rm H}C(b,D)$. When $b$ and $t$ have the same prime divisors, under certain restrictions on the digit set $D$, we give a sufficient condition for the Hausdorff measure of $W_{t}(ψ)\cap C(b,D)$ to be zero. This closes a gap in a result of Li, Li and Wu \cite{LLW2025} and shows that the dimension of the intersection can be strictly less than the product of the dimensions. The latter disproves the product conjecture of Li, Li and Wu.

math.NT↗

arXiv2025

MMhops-R1: Multimodal Multi-hop Reasoning

The ability to perform multi-modal multi-hop reasoning by iteratively integrating information across various modalities and external knowledge is critical for addressing complex real-world challenges. However, existing Multi-modal Large Language Models (MLLMs) are predominantly limited to single-step reasoning, as existing benchmarks lack the complexity needed to evaluate and drive multi-hop abilities. To bridge this gap, we introduce MMhops, a novel, large-scale benchmark designed to systematically evaluate and foster multi-modal multi-hop reasoning. MMhops dataset comprises two challenging task formats, Bridging and Comparison, which necessitate that models dynamically construct complex reasoning chains by integrating external knowledge. To tackle the challenges posed by MMhops, we propose MMhops-R1, a novel multi-modal Retrieval-Augmented Generation (mRAG) framework for dynamic reasoning. Our framework utilizes reinforcement learning to optimize the model for autonomously planning reasoning paths, formulating targeted queries, and synthesizing multi-level information. Comprehensive experiments demonstrate that MMhops-R1 significantly outperforms strong baselines on MMhops, highlighting that dynamic planning and multi-modal knowledge integration are crucial for complex reasoning. Moreover, MMhops-R1 demonstrates strong generalization to tasks requiring fixed-hop reasoning, underscoring the robustness of our dynamic planning approach. In conclusion, our work contributes a challenging new benchmark and a powerful baseline model, and we will release the associated code, data, and weights to catalyze future research in this critical area.

cs.CV↗

arXiv2025

Dimension theory of inhomogeneous Diophantine approximation with matrix sequences

In this paper, we investigate the Hausdorff dimension of naturally occurring sets of inhomogeneous well-approximable points with a sequence of real invertible matrices $\mathcal{A}=(A_n)_{n\in\mathbb{N}}$. Specifically, for a given point $\mathbf{y}\in [0,1)^d$ and a function $ψ: \mathbb{N} \to \mathbb{R}^+$, we study the limsup set \[ W\big(\mathcal{A},ψ,{\bf y}\big) =\Big\{\mathbf{x}\in [0,1)^d\colon A_n\mathbf{x}~(\bmod~1)\in B\big(\mathbf{y}, ψ(n)\big) {\rm ~ for~ infinitely ~many}~n\in\mathbb{N}\Big\}.\] The upper and lower bounds on the Hausdorff dimension of $W\big(\mathcal{A},ψ,{\bf y}\big)$ are determined by involving the singular values of $A_n$ and the successive minima of the lattice $A_n^{-1}\mathbb{Z}^d$, and both bounds are shown to be attainable for some matrices. Within this framework, we unify the problem of shrinking target sets and recurrence sets, establishing the Hausdorff dimensions for such limsup sets. As applications, our corresponding upper bounds for shrinking target and recurrence sets essentially improve those appearing in the present literature. Furthermore, explicit Hausdorff dimension formulas are derived for shrinking targets and recurrence sets associated with concrete classes of matrices. We extend the Mass Transference Principle for rectangles of Li-Liao-Velani-Wang-Zorin (Adv. Math., 2025) to rectangles under local isometries. This generalization yields a general lower bound for the Hausdorff dimension of $W\big(\mathcal{A},ψ,{\bf y}\big)$.

math.NT↗

arXiv2025

Advancing Identification method of Gamma-Ray Bursts with Data and Feature Enhancement

Gamma-ray bursts (GRBs) are challenging to identify due to their transient nature, complex temporal profiles, and limited observational datasets. We address this with a one-dimensional convolutional neural network integrated with an Adaptive Frequency Feature Enhancement module and physics-informed data augmentation. Our framework generates 100,000 synthetic GRB samples, expanding training data diversity and volume while preserving physical fidelity-especially for low-significance events. The model achieves 97.46% classification accuracy, outperforming all tested variants with conventional enhancement modules, highlighting enhanced domain-specific feature capture. Feature visualization shows model focuses on deep-seated morphological features and confirms the capability of extracting physically meaningful burst characteristics. Dimensionality reduction and clustering reveal GRBs with similar morphologies or progenitor origins cluster in the feature space, linking learned features to physical properties. This perhaps offers a novel diagnostic tool for identifying kilonova- and supernova-associated GRB candidates, establishing criteria to enhance multi-messenger early-warning systems. The framework aids current time-domain surveys, generalizes to other rare transients, and advances automated detection in large-volume observational data.

astro-ph.HE↗